int* readtree (struct aln_tree_node* p, int* tree)
{
if (p->links[0])
{
tree = readtree (p->links[0], tree);
}
if (p->links[1])
{
tree = readtree (p->links[1], tree);
}
if (p->links[0])
{
if (p->links[1])
{
tree[tree[0]] = p->links[0]->num;
tree[tree[0] + 1] = p->links[1]->num;
tree[tree[0] + 2] = p->num;
tree[0] += 3;
free (p->links[0]->internal_lables);
free (p->links[0]->links);
free (p->links[0]);
free (p->links[1]->internal_lables);
free (p->links[1]->links);
free (p->links[1]);
}
}
return tree;
}
int main() {
tree *t = readtree(0);
printtree(t), puts("");
// compute(t);
print_computed(t), puts("");
return 0;
}
void main()
{int n;float**m;
bbb*a0;
clrscr();
input(m,&n);
creattree(a0,m,n);
readtree(a0,m,n);
output(m,n);
getch();
}
// binary tree only
tree* readtree(short d) {
char *w = malloc(4096), *r, *ww = w, c;
unsigned short p = 0;
tree *t = malloc(sizeof(tree));
t->r = t->l = t->rep = 0, t->level = d, t->isrep = false;
bool f = true;
while (f)
switch (c = getchar()) {
case ')':
ungetc(')', stdin);
if (--d < 0) err(1, "unbalanced parenthesis");
f = false; break;
case ',':
ungetc(',', stdin);
f = false; break;
case EOF: f = false; break;
case '(':
t->l = readtree(d + 1);
if (*w == '\\') t->l->isrep = true;
if (getchar() != ',') err(1, "comma expected");
t->r = readtree(d + 1);
if (getchar() != ')') err(1, "closing parenthesis expected");
f = false; break;
default : if (!isspace(c)) w[p++] = c;
}
if (p) {
while (isspace(*w) && p) ++w, --p;
if (!p) return 0;
w[p] = 0;
while (isspace(w[--p])) w[p] = -2;
t->t = strdup(w);
// if (var && !strcmp(t->t, var->t)) t->rep = var;
} else {
free(t);
t = 0;
}
free(ww);
return t;
}
int main()
{
int t;
scanf("%d", &t);
for (int i = 1; i <= t; ++i) {
printf("Case #%d:\n", i);
readtree();
do_dp0();
do_dp1();
do_dp2();
do_dp3();
collect();
}
}
int main(){
value val[TRAIN_N];
FILE* ftree=fopen("tree", "rb");
FILE* fval=fopen("val", "rb");
dicisiontree tree;
FILE* finfo=fopen("fsubinfo", "rb");
float subinfo[MAX_ATTR_NUM][MAX_ATTR_VAL];
readsubinfo(subinfo, finfo);
fclose(finfo);
readtree(&tree, ftree);
readvalb(val, fval, TRAIN_N);
makesubtree(subinfo, val, &tree);
fclose(ftree);
fclose(fval);
ftree=fopen("tree", "wb");
savetree(&tree, ftree);
fclose(ftree);
return 0;
}
void profile_alignment_main(struct alignment* aln,struct parameters* param,float** submatrix)
{
float** dm = 0;
int* tree = 0;
struct aln_tree_node* tree2 = 0;
int i,j;
int tmp_numseq;
int tmp_numprofiles;
local_numseq = 0;
local_numprofiles = 0;
//determine number of profiles that were inputted....
while(aln->sl[local_numseq+numseq]){
local_numseq++;
}
local_numprofiles = (local_numseq << 1) - 1;
//fprintf(stderr,"%d %d\n",local_numseq,local_numprofiles);
for (i = 0;i < numseq;i++){
// fprintf(stderr,"%d %d %d\n",i,aln->s[i][0],aln->s[i][1]);
aln->s[i] = assign_gap_codes(aln->s[i],aln->sl[i]);
}
if(param->dna == 1){
if(byg_start(param->tree,"njNJ") != -1){
dm = dna_profile_distance(aln,dm,param,1);
}else{
dm = dna_profile_distance(aln,dm,param,0);
}
}else{
if(byg_start(param->tree,"njNJ") != -1){
dm = protein_profile_wu_distance(aln,dm,param,1);
}else{
dm = protein_profile_wu_distance(aln,dm,param,0);
}
}
/*for ( i=0; i < local_numseq;i++){
for (j = 0;j < local_numseq;j++){
fprintf(stderr,"%f ",dm[i][j]);
}
fprintf(stderr,"\n");
}*/
tmp_numseq = numseq;
tmp_numprofiles = numprofiles;
numseq = local_numseq;
numprofiles = local_numprofiles;
if(byg_start(param->tree,"njNJ") != -1){
tree2 = real_nj(dm,param->ntree);
}else{
tree2 = real_upgma(dm,param->ntree);
}
//WAs here need too add tree2 -> treee.....
tree = malloc(sizeof(int)*(numseq*3+1));
for ( i = 1; i < (numseq*3)+1;i++){
tree[i] = 0;
}
tree[0] = 1;
tree = readtree(tree2,tree);
for (i = 0; i < (numseq*3);i++){
tree[i] = tree[i+1]+ tmp_numseq;
}
//exit(0);
numseq = tmp_numseq;
numprofiles = tmp_numprofiles;
int** map = 0;
map = hirschberg_profile_alignment(aln,tree,submatrix, map);
//clear up sequence array to be reused as gap array....
int *p = 0;
for (i = 0; i < numseq;i++){
p = aln->s[i];
for (j = 0; j < aln->sl[i];j++){
p[j] = 0;
}
}
//clear up
int a,b,c;
for (i = 0; i < (local_numseq-1)*3;i +=3){
a = tree[i];
b = tree[i+1];
c = tree[i+2];
aln = make_seq(aln,a,b,map[c]);
}
for (i = 0; i < numseq;i++){
aln->nsip[i] = 0;
}
aln = sort_sequences(aln,tree,param->sort);
//for (i = 0; i < numseq;i++){
// fprintf(stderr,"%d %d %d\n",i,aln->nsip[i],aln->sip[i][0]);
//}
output(aln,param);
free(tree2->links);
free(tree2->internal_lables);
free(tree2);
free(map);
free(tree);
exit(0);
}
void KalignAdapter::alignUnsafe(const MultipleSequenceAlignment& ma, MultipleSequenceAlignment& res, TaskStateInfo& ti) {
ti.progress = 0;
int* tree = 0;
quint32 a, b, c;
struct alignment* aln = 0;
struct parameters* param = 0;
struct aln_tree_node* tree2 = 0;
param = (parameters*)malloc(sizeof(struct parameters));
param = interface(param,0,0);
kalign_context *ctx = get_kalign_context();
unsigned int &numseq = ctx->numseq;
unsigned int &numprofiles = ctx->numprofiles;
if (ma->getNumRows() < 2){
if (!numseq){
k_printf("No sequences found.\n\n");
} else {
k_printf("Only one sequence found.\n\n");
}
free_param(param);
throw KalignException("Can't align less then 2 sequences");
}
if(ctx->gpo != -1) {
param->gpo = ctx->gpo;
}
if(ctx->gpe != -1) {
param->gpe = ctx->gpe;
}
if(ctx->tgpe != -1) {
param->tgpe = ctx->tgpe;
}
if(ctx->secret != -1) {
param->secret = ctx->secret;
}
/************************************************************************/
/* Convert MA to aln */
/************************************************************************/
k_printf("Prepare data");
numseq = ma->getNumRows();
numprofiles = (numseq << 1) - 1;
aln = aln_alloc(aln);
for(quint32 i = 0 ; i < numseq; i++) {
const MultipleSequenceAlignmentRow row= ma->getMsaRow(i);
aln->sl[i] = row->getUngappedLength();
aln->lsn[i] = row->getName().length();
}
for (quint32 i = 0; i < numseq;i++) {
try {
aln->s[i] = (int*) malloc(sizeof(int)*(aln->sl[i]+1));
checkAllocatedMemory(aln->s[i]);
aln->seq[i] = (char*)malloc(sizeof(char)*(aln->sl[i]+1));
checkAllocatedMemory(aln->seq[i]);
aln->sn[i] = (char*)malloc(sizeof(char)*(aln->lsn[i]+1));
checkAllocatedMemory(aln->sn[i]);
} catch (...) {
cleanupMemory(NULL, numseq, NULL, aln, param);
throw;
}
}
int aacode[26] = {0,1,2,3,4,5,6,7,8,-1,9,10,11,12,23,13,14,15,16,17,17,18,19,20,21,22};
for(quint32 i = 0; i < numseq; i++) {
const MultipleSequenceAlignmentRow row= ma->getMsaRow(i);
qstrncpy(aln->sn[i], row->getName().toLatin1(), row->getName().length() + 1); //+1 to include '\0'
QString gapless = QString(row->getCore()).remove('-');
qstrncpy(aln->seq[i], gapless.toLatin1(), gapless.length() + 1); //+1 to include '\0'
for (quint32 j = 0; j < aln->sl[i]; j++) {
if (isalpha((int)aln->seq[i][j])){
aln->s[i][j] = aacode[toupper(aln->seq[i][j])-65];
} else {
aln->s[i][j] = -1;
}
}
aln->s[i][aln->sl[i]] = 0;
aln->seq[i][aln->sl[i]] = 0;
aln->sn[i][aln->lsn[i]] = 0;
}
/*for(int i=0;i<numseq;i++) {
for(int j=0;j<aln->sl[i];j++)
printf("%d ", aln->s[i][j]);
}*/
//aln_dump(aln);
//aln = detect_and_read_sequences(aln,param);
if(param->ntree > (int)numseq){
param->ntree = (int)numseq;
}
//DETECT DNA
if(param->dna == -1){
for (quint32 i = 0; i < numseq;i++){
param->dna = byg_detect(aln->s[i],aln->sl[i]);
if(param->dna){
break;
}
}
}
//param->dna = 0;
//k_printf("DNA:%d\n",param->dna);
//exit(0);
if(param->dna == 1){
//brief sanity check...
for (quint32 i = 0; i < numseq;i++){
if(aln->sl[i] < 6){
//k_printf("Dna/Rna alignments are only supported for sequences longer than 6.");
free(param);
free_aln(aln);
throw KalignException("Dna/Rna alignments are only supported for sequences longer than 6.");
}
}
aln = make_dna(aln);
}
//int j;
//fast distance calculation;
float** submatrix = 0;
submatrix = read_matrix(submatrix,param); // sets gap penalties as well.....
//if(byg_start(param->alignment_type,"profPROFprofilePROFILE") != -1){
// profile_alignment_main(aln,param,submatrix);
//}
float** dm = 0;
if(param->ntree > 1){
//if(byg_start(param->distance,"pairclustalPAIRCLUSTAL") != -1){
// if(byg_start(param->tree,"njNJ") != -1){
// dm = protein_pairwise_alignment_distance(aln,dm,param,submatrix,1);
// }else{
// dm = protein_pairwise_alignment_distance(aln,dm,param,submatrix,0);
// }
//}else if(byg_start("wu",param->alignment_type) != -1){
// dm = protein_wu_distance2(aln,dm,param);
// // param->feature_type = "wumanber";
if(param->dna == 1){
// if(byg_start(param->tree,"njNJ") != -1){
// dm = dna_distance(aln,dm,param,1);
// }else{
dm = dna_distance(aln,dm,param,0);
// }
}else{
//if(byg_start(param->tree,"njNJ") != -1){
// dm = protein_wu_distance(aln,dm,param,1);
//}else{
try {
dm = protein_wu_distance(aln,dm,param,0);
} catch (const KalignException &) {
cleanupMemory(submatrix, numseq, dm, aln, param);
throw;
}
//}
}
if(check_task_canceled(ctx)) {
cleanupMemory(submatrix, numseq, dm, aln, param);
throwCancellingException();
}
/*int j;
for (int i = 0; i< numseq;i++){
for (j = 0; j< numseq;j++){
k_printf("%f ",dm[i][j]);
}
k_printf("\n");
}*/
//if(byg_start(param->tree,"njNJ") != -1){
// tree2 = real_nj(dm,param->ntree);
//}else{
tree2 = real_upgma(dm,param->ntree);
//}
//if(param->print_tree){
// print_tree(tree2,aln,param->print_tree);
//}
}
tree = (int*) malloc(sizeof(int)*(numseq*3+1));
for (quint32 i = 1; i < (numseq*3)+1;i++){
tree[i] = 0;
}
tree[0] = 1;
if(param->ntree < 2){
tree[0] = 0;
tree[1] = 1;
c = numseq;
tree[2] = c;
a = 2;
for (quint32 i = 3; i < (numseq-1)*3;i+=3){
tree[i] = c;
tree[i+1] = a;
c++;
tree[i+2] = c;
a++;
}
}else if(param->ntree > 2){
ntreeify(tree2,param->ntree);
}else{
tree = readtree(tree2,tree);
for (quint32 i = 0; i < (numseq*3);i++){
tree[i] = tree[i+1];
}
free(tree2->links);
free(tree2->internal_lables);
free(tree2);
}
//get matrices...
//struct feature_matrix* fm = 0;
//struct ntree_data* ntree_data = 0;
int** map = 0;
//if(param->ntree > 2){
// ntree_data = (struct ntree_data*)malloc(sizeof(struct ntree_data));
// ntree_data->realtree = tree2;
// ntree_data->aln = aln;
// ntree_data->profile = 0;
// ntree_data->map = 0;
// ntree_data->ntree = param->ntree;
// ntree_data->submatrix = submatrix;
// ntree_data->tree = tree;
// ntree_data = ntree_alignment(ntree_data);
// map = ntree_data->map;
// tree = ntree_data->tree;
// for (int i = 0; i < (numseq*3);i++){
// tree[i] = tree[i+1];
// }
// free(ntree_data);
//}else if (param->feature_type){
// fm = get_feature_matrix(fm,aln,param);
// if(!fm){
// for (int i = 32;i--;){
// free(submatrix[i]);
// }
// free(submatrix);
// free_param(param);
// free(map);
// free(tree);
// throw KalignException("getting feature matrix error");
// }
// map = feature_hirschberg_alignment(aln,tree,submatrix,map,fm);
// //exit(0);
// //map = feature_alignment(aln,tree,submatrix, map,fm);
//}else if (byg_start("pairwise",param->alignment_type) != -1){
// if(param->dna == 1){
// map = dna_alignment_against_a(aln,tree,submatrix, map,param->gap_inc);
// }else{
// map = hirschberg_alignment_against_a(aln,tree,submatrix, map,param->smooth_window,param->gap_inc);
// }
// //map = default_alignment(aln,tree,submatrix, map);
//}else if (byg_start("fast",param->alignment_type) != -1){
// map = default_alignment(aln,tree,submatrix, map);
if(param->dna == 1){
map = dna_alignment(aln,tree,submatrix, map,param->gap_inc);
// /*}else if (byg_start("test",param->alignment_type) != -1){
// map = test_alignment(aln,tree,submatrix, map,param->internal_gap_weight,param->smooth_window,param->gap_inc);
// }else if (param->aa){
// map = aa_alignment(aln,tree,submatrix, map,param->aa);
// }else if (param->alter_gaps){
// map = alter_gaps_alignment(aln,tree,submatrix,map,param->alter_gaps,param->alter_range,param->alter_weight);
// }else if (byg_start("altergaps",param->alignment_type) != -1){
// map = alter_gaps_alignment(aln,tree,submatrix,map,param->alter_gaps,param->alter_range,param->alter_weight);
// }else if(byg_start("simple",param->alignment_type) != -1){
// map = simple_hirschberg_alignment(aln,tree,submatrix, map);*/
//}else if(byg_start("advanced",param->alignment_type) != -1){
// map = advanced_hirschberg_alignment(aln,tree,submatrix, map,param->smooth_window,param->gap_inc,param->internal_gap_weight);
}else{
map = hirschberg_alignment(aln,tree,submatrix, map,param->smooth_window,param->gap_inc);
}
if (map == NULL) {
throw KalignException("Failed to build alignment.");
}
if(check_task_canceled(ctx)) {
free_param(param);
free_aln(aln);
free(map);
free(tree);
throwCancellingException();
}
//clear up sequence array to be reused as gap array....
int *p = 0;
for (quint32 i = 0; i < numseq;i++){
p = aln->s[i];
for (a = 0; a < aln->sl[i];a++){
p[a] = 0;
}
}
//clear up
for (quint32 i = 0; i < (numseq-1)*3;i +=3){
a = tree[i];
b = tree[i+1];
aln = make_seq(aln,a,b,map[tree[i+2]]);
}
//for (int i = 0; i < numseq;i++){
// k_printf("%s %d\n",aln->sn[i],aln->nsip[i]);
//}
for (quint32 i = 0; i < numseq;i++){
aln->nsip[i] = 0;
}
aln = sort_sequences(aln,tree,param->sort);
//for (int i = 0; i < numseq;i++){
// k_printf("%d %d %d\n",i,aln->nsip[i],aln->sip[i][0]);
//}
/************************************************************************/
/* Convert aln to MA */
/************************************************************************/
res->setAlphabet(ma->getAlphabet());
for (quint32 i = 0; i < numseq;i++){
int f = aln->nsip[i];
QString seq;
for(quint32 j=0;j<aln->sl[f];j++) {
seq += QString(aln->s[f][j],'-') + aln->seq[f][j];
}
seq += QString(aln->s[f][aln->sl[f]],'-');
res->addRow(QString(aln->sn[f]), seq.toLatin1());
}
//output(aln,param);
/* if(!param->format){
fasta_output(aln,param->outfile);
}else{
if (byg_start("msf",param->format) != -1){
msf_output(aln,param->outfile);
}else if (byg_start("clustal",param->format) != -1){
clustal_output(aln,param->outfile);
}else if (byg_start("macsim",param->format) != -1){
macsim_output(aln,param->outfile,param->infile[0]);
}
}
*/
free_param(param);
free_aln(aln);
free(map);
free(tree);
//KalignContext* ctx = getKalignContext();
}
int main(int argc, char** argv){
/* A->0 ; C->1 ; G->2 ; T->3 */
int i, j, nbseq, nb1, nb2, prov, secondes, minutes, heures, jours, muet, nbeval;
char nomfinseq[50], nomfintree[50], nomfopt[50], *seq[MAXNSP], *seqname[MAXNSP], *comments[MAXNSP];
char *ctree, *current_best, **evaluated, *ret;
double maxl, runtime;
FILE *treefile,*in, *optfile, *current_best_file, *evaluated_file;
options opt;
time_t debut, fin;
/* debugf=fopen("shake_debug", "w");*/
init_cas();
srand48(seed);
/** input sequences **/
if(argc<2){
muet=0;
while(1){
printf("\nSequence file (MASE) ? ");
gets(nomfinseq);
in=fopen(nomfinseq, "r");
if(in) break;
printf("Cannot find file : %s\n", nomfinseq);
}
}
else{
in=fopen(argv[1], "r");
if(!in){
printf("Cannot find sequence file : %s\n", argv[1]);
exit(EXIT_FAILURE);
}
sprintf(nomfinseq, "%s", argv[1]);
muet=1;
}
fclose(in);
if(muet)
nbseq=readmasemuet(nomfinseq, seq, seqname, comments, MAXDATASET*MAXNSP);
else
nbseq=readmaseseqs(nomfinseq, seq, seqname, comments, MAXDATASET*MAXNSP);
refresh(seq, nbseq, 0);
/* debug_nom=seqname; */
/** input tree **/
if(argc<3){
while(1){
printf("\nTree file ? ");
gets(nomfintree);
treefile=fopen(nomfintree, "r");
if(treefile) break;
printf("Cannot find file : %s\n", nomfintree);
}
}
else{
treefile=fopen(argv[2], "r");
if(!treefile){
printf("Cannot find tree file : %s\n", argv[2]);
exit(EXIT_FAILURE);
}
}
ctree=readtree(treefile, nbseq);
/** input options **/
if(argc<4){
while(1){
printf("\nOption file ? ");
gets(nomfopt);
optfile=fopen(nomfopt, "r");
if(optfile) break;
printf("Cannot find file : %s\n", nomfopt);
}
}
else{
optfile=fopen(argv[3], "r");
if(!optfile){
printf("Cannot find option file : %s\n", argv[3]);
exit(EXIT_FAILURE);
}
}
getoptions(&opt, optfile);
/** input saved work **/
if(opt->SH_RESTART && argc==5){
evaluated_file=fopen(argv[4], "r");
if(!evaluated_file){
printf("Cannot find evaluated file %s\n", argv[5]);
exit(EXIT_FAILURE);
}
evaluated=check_alloc(MAX_NBLISTMAX, sizeof(char*));
nbeval=0; j=3*MAXLNAME*nbseq;
while(1){
evaluated[nbeval]=check_alloc(j, sizeof(char));
ret=fgets(evaluated[nbeval], j, evaluated_file);
if(ret==NULL) break;
nbeval++;
}
}
else{
evaluated=NULL; nbeval=0;
}
time(&debut);
/** shake **/
maxl=shake(nbseq, seq, seqname, ctree, opt, evaluated, nbeval);
/** output **/
if(!opt->print->PRINT1)
printf("%f\n", maxl);
time(&fin);
runtime=difftime(fin, debut);
if(opt->print->PRINT1){
jours=(int)(runtime/86400.);
heures=(int)((runtime-86400*jours)/3600.);
minutes=(int)((runtime-86400*jours-3600*heures)/60.);
secondes=(int)(runtime-86400*jours-3600*heures-60*minutes);
printf("(running time : ");
if(jours) printf("%d days, ", jours);
if(jours || heures) printf("%d hours, ", heures);
if(jours || heures || minutes) printf("%d minutes, ", minutes);
printf("%d seconds)\n\n", secondes);
}
return 0;
}